******************************************************************
******************************************************************
*****                                                        *****
*****       Mona Morgan-Collins (King's College London)      *****
*****        Contact: mona.morgan-collins@dkcl.ac.uk         *****
*****                                                        *****
*****     (Letter) How Gap Measures Determine Results:       *****       
*****    The Case of Proportional Systems and the Gender     ***** 
*****                 Mobilization Gap                       *****
*****                                                        *****
*****        British Journal of Political Science            *****
*****                                                        *****
*****       Replicating Analyses in the Appendix             *****
*****                                                        *****
******************************************************************
****************************************************************** 

*set seed  - setting the initial value of the random-number seed set as Stata's default when Stata is launched.
set seed 123456789
/* Because bootstrapping involves drawing pseudorandom numbers, the exact results of BOOTTEST command */
/* depend on the starting value of the random-number generator. */
/* This means that all  wild bootstrap results can be reproduced only on the first run of each do file */
/* after Stata is launched and when all BOOTTEST commands are run consecutively in each do file */ 
/* OR when seed is set as above before each do file is run and when all BOOTTEST commands are run consecutively in each do file */ 

*****************************
*Figure A1: Predicting Sensitivity of Measures to Electoral Context, Alternative Assumptions
*****************************
*Nothing to replicate


*****************************
*Table A1: Variable Definitions
*****************************
*Nothing to replicate


*****************************
*Table A2: Summary Statistics
*****************************
*LOCAL 
use dta\munic1619bal, clear                              //using munic1619bal data set                  
sum turnoutw   if  maj==1 & year==1916 
sum turnoutm   if  maj==1 & year==1916 
sum turng      if  maj==1 & year==1916 
sum turnoutwsh if  maj==1 & year==1916 

sum turnoutw   if  maj16==1 & year==1919 
sum turnoutm   if  maj16==1 & year==1919 
sum turng      if  maj16==1 & year==1919 
sum turnoutwsh if  maj16==1 & year==1919 

sum ch_turnoutw   if  maj16==1 & year==1919 
sum ch_turnoutm   if  maj16==1 & year==1919 
sum ch_turnoutg   if  maj16==1 & year==1919 
sum ch_turnoutwsh if  maj16==1 & year==1919 

*PARL
use dta\parl, clear                              //using parl data set
sum turnoutw   if  year==1918 
sum turnoutm   if  year==1918 
sum turng      if  year==1918 
sum turnoutwsh if  year==1918 

sum turnoutw   if  year==1921 
sum turnoutm   if  year==1921 
sum turng      if  year==1921 
sum turnoutwsh if  year==1921 

sum ch_turnoutw   if  year==1921 
sum ch_turnoutm   if  year==1921 
sum ch_turnoutg   if  year==1921
sum ch_turnoutwsh if  year==1921


*****************************
*Table A3: The Effect of Men's Turnout on Gender Gap Measures 
*****************************
*Model 1: Local 1916
use dta\munic1619bal, clear                              //using munic1619bal data set
reg turnoutwsh  turnoutm  fact20 intel20 marw20 i.urban     , robust  ,  if maj==1 & year==1916
*Model 2: Parl 1918 
use dta\parl, clear                                   //using parl data set
reg turnoutwsh  turnoutm  fact20 intel20 marw20 i.urban   i.PR_district    , cluster(PR_district)  ,  if year==1918
boottest turnoutm
*Model 3: Local 1916
use dta\munic1619bal, clear                              //using munic1619bal data set
reg turng   c.turnoutm##c.turnoutm   fact20 intel20 marw20 i.urban     , robust  ,  if maj==1 & year==1916
*Model 4: Parl 1918 
use dta\parl, clear                                   //using parl data set
reg turng  c.turnoutm##c.turnoutm  fact20 intel20 marw20 i.urban   i.PR_district    , cluster(PR_district)  ,  if year==1918
gen turnoutm_sq = turnoutm*turnoutm
reg turng  turnoutm turnoutm_sq  fact20 intel20 marw20 i.urban   i.PR_district    , cluster(PR_district)  ,  if year==1918
boottest turnoutm
boottest turnoutm_sq


*****************************
*Figure A2: The Effect of Men's Turnout on Gender Gap Measures, Adjusted Means
*****************************
** sub-figure a
*Local 1916 
use dta\munic1619bal, clear                              //using munic1619bal data set
reg turnoutwsh  turnoutm  fact20 intel20 marw20 i.urban     , robust  ,  if maj==1 & year==1916
margins, at(turnoutm = (9 (1) 91))
#delimit ;
 marginsplot, 
 title("Local 1916", size(vlarge) ) 
 ytitle("Linear Prediction", size(vlarge) ) 
 ylabel(0(20)60, labsize(vlarge) ) 
 xtitle ("Men's Turnout", size(vlarge)) 
 xlabel(0(25)100, labsize(vlarge) ) 
 plotopts(msymbol(o) msize(huge)  lwidth(thick) lpattern(solid) color(black)) 
 ciopts(lpattern(dash) color(black) lwidth(medthick) msymbol (none)) recast(line) recastci(rconnected)  
 legend(off) scheme(s1mono) ysize(6)   ;
#delimit cr
*Parl 1918
use dta\parl, clear
reg turnoutwsh  turnoutm  fact20 intel20 marw20 i.urban  i.PR_district    , cluster(PR_district)  ,  if year==1918  
margins, at(turnoutm = (38 (1) 91))
#delimit ;
 marginsplot, 
 title("Parliamentary 1918", size(vlarge) ) 
 ytitle("Linear Prediction", size(vlarge) ) 
 ylabel(0(20)60, labsize(vlarge) ) 
 xtitle ("Men's Turnout", size(vlarge)) 
 xlabel(0(25)100, labsize(vlarge) ) 
 plotopts(msymbol(o) msize(huge)  lwidth(thick) lpattern(solid) color(black)) 
 ciopts(lpattern(dash) color(black) lwidth(medthick) msymbol (none)) recast(line) recastci(rconnected)  
 legend(off) scheme(s1mono) ysize(6)   ;
#delimit cr
** sub-figure b
*Local 1916 
use dta\munic1619bal, clear                              //using munic1619bal data set
reg turng   c.turnoutm##c.turnoutm   fact20 intel20 marw20 i.urban     , robust  ,  if maj==1 & year==1916
margins, at(turnoutm = (9 (1) 91))
#delimit ;
 marginsplot, 
 title("Local 1916", size(vlarge) ) 
 ytitle("Linear Prediction", size(vlarge) ) 
 ylabel(-50(25)25, labsize(vlarge) ) 
 xtitle ("Men's Turnout", size(vlarge)) 
 xlabel(0(25)100, labsize(vlarge) ) 
 plotopts(msymbol(o) msize(huge)  lwidth(thick) lpattern(solid) color(black)) 
 ciopts(lpattern(dash) color(black) lwidth(medthick) msymbol (none)) recast(line) recastci(rconnected)  
 yline(0, lcolor(black)) legend(off) scheme(s1mono) ysize(6)   ;
#delimit cr
*Parl 1918 
use dta\parl, clear
reg turng  c.turnoutm##c.turnoutm  fact20 intel20 marw20 i.urban  i.PR_district    , cluster(PR_district)  ,  if year==1918  
margins, at(turnoutm = (38 (1) 91)) 
#delimit ;
 marginsplot, 
 title("Parliamentary 1918", size(vlarge) ) 
 ytitle("Linear Prediction", size(vlarge) ) 
 ylabel(-50(25)25, labsize(vlarge) ) 
 xtitle ("Men's Turnout", size(vlarge)) 
 xlabel(0(25)100, labsize(vlarge) ) 
 plotopts(msymbol(o) msize(huge)  lwidth(thick) lpattern(solid) color(black)) 
 ciopts(lpattern(dash) color(black) lwidth(medthick) msymbol (none)) recast(line) recastci(rconnected)  
 yline(0, lcolor(black)) legend(off) scheme(s1mono) ysize(6)   ;
#delimit cr


*****************************
*Figure A3: Kernel Density of Women's and Men's Pre-Reform Turnout
*****************************
*sub-figure a
use dta\parl, clear                              //using parl data set
#delimit ; 
twoway
(kdensity  turnoutm if  year==1918, lcolor(blue) lwidth(thick) )
(kdensity  turnoutw if  year==1918, lcolor(red) lwidth(thick)),
 legend(off)  scheme(s1mono) title("",  size(huge)) ysize(5)
 xlabel(0(25)100, labsize(vlarge)) xtitle("Pre-Reform Turnout (1918, SMD)", size(vlarge)) 
 ylabel(0(0.025)0.1, labsize(vlarge)) ytitle("Density", size(vlarge))  
 text(0.095 80 "Women's Turnout", color(red) size(med)) 
 text(0.09 76 "Men's Turnout", color(blue) size(med)) ; 
#delimit cr
*sub-figure b 
use dta\munic1619bal, clear                              //using munic1619bal data set
#delimit ; 
twoway
(kdensity  turnoutm if maj==1 & year==1916, lcolor(blue) lwidth(thick))
(kdensity  turnoutw if maj==1 & year==1916, lcolor(red) lwidth(thick)),
 legend(off)  scheme(s1mono) title("",  size(huge)) ysize(5)
 xlabel(0(25)100, labsize(vlarge)) xtitle("Pre-Reform Turnout (1916, SMD)", size(vlarge)) 
 ylabel(0(0.025)0.1, labsize(vlarge)) ytitle("Density", size(vlarge))  
 text(0.095 80 "Women's Turnout", color(red) size(med)) 
 text(0.09 76 "Men's Turnout", color(blue) size(med)) ; 
#delimit cr


*****************************
*Table A4: The Effect of Pre-Reform Men's Turnout on Change in Gap Measures 
*****************************
use dta\munic1619bal, clear                              //using munic1619bal data set
*Model 1
reg ch_turnoutwsh  turnoutm16  fact20 intel20 marw20 i.urban     , robust  ,  if maj16==1 & year==1919  
*Model 3
reg ch_turnoutg  c.turnoutm16##c.turnoutm16   fact20 intel20 marw20 i.urban    , robust  ,  if maj16==1 & year==1919  
use dta\parl, clear                              //using parl data set
*Model 2
reg  ch_turnoutwsh Lturnoutm fact20 intel20 marw20 i.urban   i.PR_district   if  year==1921, cluster(PR_district)
boottest Lturnoutm
*Model 4
reg  ch_turnoutg c.Lturnoutm##c.Lturnoutm fact20 intel20 marw20 i.urban  i.PR_district if  year==1921, cluster(PR_district)
gen Lturnoutm_sq = Lturnoutm*Lturnoutm
reg  ch_turnoutg Lturnoutm Lturnoutm_sq fact20 intel20 marw20 i.urban  i.PR_district if  year==1921, cluster(PR_district)
boottest Lturnoutm
boottest Lturnoutm_sq


*****************************
*Table A5: The Effect of PR on Gap Measures (Difference-in-Differences)
*****************************
use dta\munic1619bal, clear                              //using munic1619bal data set
*Model 1
reg turnoutwsh post treat did fact20 intel20 marw20 i.urban  , r
*Model 2
reg turng post treat did fact20 intel20 marw20 i.urban  , r

use dta\munic1319bal, clear                              //using munic1319bal data set
*Model 3
reg turnoutwsh post treat did fact20 intel20 marw20 i.urban  , r, if yr1619==1 
*Model 4
reg turng post treat did fact20 intel20 marw20 i.urban  , r, if yr1619==1 
*Model 5
reg turnoutwsh post_pl treat_pl did_pl  fact20 intel20 marw20 i.urban  , r, if yr1316==1 
*Model 6
reg turng post_pl treat_pl did_pl  fact20 intel20 marw20 i.urban  , r, if yr1316==1 


*****************************
*Table A6: The Heterogeneous Effect of PR by Pre-Reform Men's Turnout (Difference-in-Differences)
*****************************
use dta\munic1619bal, clear                              //using munic1619bal data set

*Model 1
*reg turng i.post##i.treat##c.turnoutm16 fact20 intel20 marw20 i.urban  , r
reg turng i.post i.treat i.did i.post##c.turnoutm16  i.treat##c.turnoutm16 i.did##c.turnoutm16     fact20 intel20 marw20 i.urban  , r

*Model 2
*reg turng i.post##i.treat##c.turnoutm16##c.turnoutm16 fact20 intel20 marw20 i.urban  , r
reg turng i.post i.treat i.did i.post##c.turnoutm16##c.turnoutm16  i.treat##c.turnoutm16##c.turnoutm16   i.did##c.turnoutm16##c.turnoutm16       fact20 intel20 marw20 i.urban  , r


*****************************
*Figure A4: The Heterogeneous Effect of PR by Pre-Reform Men's Turnout (Difference-in-Differences)
*****************************
use dta\munic1619bal, clear                              //using munic1619bal data set

*sub-gigure a
*reg turng i.post##i.treat##c.turnoutm16 fact20 intel20 marw20 i.urban  , r
reg turng i.post i.treat i.did i.post##c.turnoutm16  i.treat##c.turnoutm16 i.did##c.turnoutm16     fact20 intel20 marw20 i.urban  , r
margins, dydx(did) at(turnoutm16=(9(1)97)  ) vsquish noestimcheck 
#delimit;
 marginsplot, addplot (hist turnoutm16,  color(none) lcolor(gray) lwidth(0.05) yaxis(2) yscale(alt) yscale(alt axis(2)) 
 ytitle("Heterogeneous Effects of PR", size(vlarge) axis(1)) ytitle("Observations (%)", size(vlarge) axis(2))
 ylabel(0(0.01)0.04, labsize(vlarge) axis(2)) ylabel(-30(15)30, labsize(vlarge) axis(1)) xlabel(, labsize(vlarge)))
 title("", size(vlarge)) xtitle ("Pre-Reform Men's Turnout", size(vlarge)) 
 plotopts(msymbol(o) msize(vlarge) lwidth(med) lpattern(solid) color()) ciopts(lpattern(dash) color() lwidth(medthick) msymbol (none)) recastci(rconnected)  
 yline(0, lcolor(black)) legend(off ) yscale(alt) scheme(s1mono) ysize(5)  ;
#delimit cr

*sub-figure b
*reg turng i.post##i.treat##c.turnoutm16##c.turnoutm16 fact20 intel20 marw20 i.urban  , r
reg turng i.post i.treat i.did i.post##c.turnoutm16##c.turnoutm16  i.treat##c.turnoutm16##c.turnoutm16   i.did##c.turnoutm16##c.turnoutm16       fact20 intel20 marw20 i.urban  , r
margins, dydx(did) at(turnoutm16=(9(1)97)  ) vsquish noestimcheck 
#delimit;
 marginsplot, addplot (hist turnoutm16,  color(none) lcolor(gray) lwidth(0.05) yaxis(2) yscale(alt) yscale(alt axis(2)) 
 ytitle("Heterogeneous Effects of PR", size(vlarge) axis(1)) ytitle("Observations (%)", size(vlarge) axis(2))
 ylabel(0(0.01)0.04, labsize(vlarge) axis(2)) ylabel(-30(15)30, labsize(vlarge) axis(1)) xlabel(, labsize(vlarge)))
 title("", size(vlarge)) xtitle ("Pre-Reform Men's Turnout", size(vlarge)) 
 plotopts(msymbol(o) msize(vlarge) lwidth(med) lpattern(solid) color()) ciopts(lpattern(dash) color() lwidth(medthick) msymbol (none)) recastci(rconnected)  
 yline(0, lcolor(black)) legend(off ) yscale(alt) scheme(s1mono) ysize(5)  ;
#delimit cr


*********************
* Figure A5: Plotting Difference-Based Measure Against Proportion-Based Measure
*********************
*sub-figure a
use dta\parl, clear                              //using parl data set
corr turng  turnoutwsh if  year==1918
local corr : di %4.3f r(rho)
#delimit ;
 twoway (scatter turng  turnoutwsh if  year==1918    , mcolor(gs13) m(oh)) 
 (lowess turng  turnoutwsh if  year==1918   , lwidth(thick) lcolor(gray) )
 (lfit turng  turnoutwsh if  year==1918   , lwidth(thick) lcolor(black) )
 , ytitle(Difference in Turnout (M-W), size(vlarge)) ylabel(-80(20)20, labsize(huge)) 
 xtitle(Women's Share Among Voters, size(vlarge)) xlabel(0(20)60, labsize(huge) ) 
 legend(off) scheme(s1mono) yline(0) ysize(5)  note(r=`corr', size(med))  ;
#delimit cr
*sub-figure b
use dta\munic1619bal, clear                              //using munic1619bal data set
corr turng  turnoutwsh if  maj==1 & year==1916
local corr : di %4.3f r(rho)
#delimit ;
 twoway (scatter turng  turnoutwsh if  maj==1 & year==1916    , mcolor(gs13) m(oh)) 
 (lowess turng  turnoutwsh if  maj==1 & year==1916   , lwidth(thick) lcolor(gray) )
 (lfit turng  turnoutwsh if  maj==1 & year==1916   , lwidth(thick) lcolor(black)  )
 , ytitle(Difference in Turnout (M-W), size(vlarge)) ylabel(-80(20)20, labsize(huge)) 
 xtitle(Women's Share Among Voters, size(vlarge)) xlabel(0(20)60, labsize(huge) ) 
 legend(off) scheme(s1mono) yline(0) ysize(5) note(r=`corr', size(med)) ;
#delimit cr


*****************************
*Table A7: Alternative Measures Used Outside of the PR-Gap Debate 
*****************************
*nothing to replicate


*****************************
*Figure A6: Kernel Density of Change in Alternative Measures
*****************************
*sub-figure a
use dta\parl, clear                              //using parl data set
*Vote Ratio
#delimit ; 
twoway(kdensity  ch_votrat if  year==1921, lcolor(black) lwidth(thick))
(kdensity  ch_votrat if year==1918, lcolor(gray) lwidth(thick) lpattern(dash))
, legend(off) xline(0, lcolor(black)) scheme(s1mono) title("",  size(huge)) ysize(5)
 xlabel(-1(0.5)1.5, labsize(vlarge)) xtitle("Change in Women-to-Men Vote Ratio", size(large)) 
 ylabel(0(1)4, labsize(vlarge)) ytitle("Density", size(vlarge)) 
 text(3.3 0.5 "Introduced PR", color(black) size(med)) 
 text(3.3 -0.5 "Did not introduce PR", color(gray) size(med));  
#delimit cr
*Turnout Ratio
#delimit ; 
twoway(kdensity  ch_turnrat if  year==1921, lcolor(black) lwidth(thick))
(kdensity  ch_turnrat if year==1918, lcolor(gray) lwidth(thick) lpattern(dash))
, legend(off) xline(0, lcolor(black)) scheme(s1mono) title("",  size(huge)) ysize(5)
 xlabel(-1(0.5)1.5, labsize(vlarge)) xtitle("Change in Women-to-Men Turnout Ratio", size(large)) 
 ylabel(0(1)4, labsize(vlarge)) ytitle("Density", size(vlarge))  
 text(3.6 0.5 "Introduced PR", color(black) size(med)) 
 text(3.6 -0.5 "Did not introduce PR", color(gray) size(med));  
#delimit cr 

*sub-figure b  
use dta\munic1619bal, clear                              //using munic1619bal data set
*Vote Ratio
#delimit ; 
twoway(kdensity  ch_votrat if maj16==1 & year==1919, lcolor(black) lwidth(thick))
(kdensity  ch_votrat if maj16==0 & year==1919, lcolor(gray) lwidth(thick) lpattern(dash)),
 legend(off) xline(0, lcolor(black)) scheme(s1mono) title("",  size(huge)) ysize(5)
 xlabel(-1(0.5)1.5, labsize(vlarge)) xtitle("Change in Women-to-Men Vote Ratio", size(large)) 
 ylabel(0(1)4, labsize(vlarge)) ytitle("Density", size(vlarge)) 
 text(1.8 0.7 "Introduced PR", color(black) size(med)) 
 text(2.8 0.6 "Did not introduce PR", color(gray) size(med)); 
#delimit cr
*Turnout Ratio
#delimit ; 
twoway(kdensity  ch_turnrat if maj16==1 & year==1919, lcolor(black) lwidth(thick))
(kdensity  ch_turnrat if maj16==0 & year==1919, lcolor(gray) lwidth(thick) lpattern(dash)),
 legend(off) xline(0, lcolor(black)) scheme(s1mono) title("",  size(huge)) ysize(5)
 xlabel(-1(0.5)1.5, labsize(vlarge)) xtitle("Change in Women-to-Men Turnout Ratio", size(large)) 
 ylabel(0(1)4, labsize(vlarge)) ytitle("Density", size(vlarge))  
 text(1.8 0.7 "Introduced PR", color(black) size(med)) 
 text(2.8 0.63 "Did not introduce PR", color(gray) size(med)); 
#delimit cr


*****************************
*Table A8: The Effect of Pre-Reform Men's Turnout on Change in Alternative Measures 
*****************************
use dta\munic1619bal, clear                              //using munic1619bal data set
*Model 1
reg ch_votrat   turnoutm16  fact20 intel20 marw20 i.urban     , robust  ,  if maj16==1 & year==1919  
*Model 3
reg ch_turnrat  turnoutm16  fact20 intel20 marw20 i.urban     , robust  ,  if maj16==1 & year==1919  

use dta\parl, clear                                   //using parl data set
*Model 2
reg  ch_votrat  Lturnoutm fact20 intel20 marw20 i.urban    i.PR_district, cluster(PR_district), if  year==1921
boottest Lturnoutm
*Model 4
reg  ch_turnrat  Lturnoutm fact20 intel20 marw20 i.urban   i.PR_district, cluster(PR_district), if  year==1921
boottest Lturnoutm

*****************************
*Figure A7: The Effect of Pre-Reform Men's Turnout on Change in Alternative Gap Measures; Adjusted Means 
*****************************
*LOCAL 
use dta\munic1619bal, clear                              //using munic1619bal data set
*subfigure a
reg ch_votrat   turnoutm16  fact20 intel20 marw20 i.urban     , robust  ,  if maj16==1 & year==1919  
margins, at(turnoutm16 = (9 (1) 91))
#delimit ;
 marginsplot, 
 title("Local Elections", size(vlarge) ) 
 ytitle("Linear Prediction", size(vlarge) ) 
 ylabel(-0.2(0.2)0.4, labsize(vlarge) ) 
 xtitle ("Pre-Reform Men's Turnout (1916)", size(vlarge)) 
 xlabel(0(25)100, labsize(vlarge) ) 
 plotopts(msymbol(o) msize(huge)  lwidth(thick) lpattern(solid) color(black)) 
 ciopts(lpattern(dash) color(black) lwidth(medthick) msymbol (none)) recast(line) recastci(rconnected)  
 legend(off) scheme(s1mono) ysize(5) yline(0)  ;
#delimit cr
*sub-figure b
reg ch_turnrat  turnoutm16  fact20 intel20 marw20 i.urban     , robust  ,  if maj16==1 & year==1919  
margins, at(turnoutm16 = (9 (1) 91))
#delimit ;
 marginsplot, 
 title("Local Elections", size(vlarge) ) 
 ytitle("Linear Prediction", size(vlarge) ) 
 ylabel(-0.2(0.2)0.4, labsize(vlarge) ) 
 xtitle ("Pre-Reform Men's Turnout (1916)", size(vlarge)) 
 xlabel(0(25)100, labsize(vlarge) ) 
 plotopts(msymbol(o) msize(huge)  lwidth(thick) lpattern(solid) color(black)) 
 ciopts(lpattern(dash) color(black) lwidth(medthick) msymbol (none)) recast(line) recastci(rconnected)  
 legend(off) scheme(s1mono) ysize(5) yline(0)  ;
#delimit cr

*PARLIAMENTARY 
use dta\parl, clear                              //using parl data set
*sub-figure a
reg  ch_votrat  Lturnoutm fact20 intel20 marw20 i.urban    i.PR_district, cluster(PR_district), if  year==1921
margins, at(Lturnoutm = (38 (1) 91))
#delimit ;
 marginsplot, 
 title("Parliamentary Elections", size(vlarge) ) 
 ytitle("Linear Prediction", size(vlarge) ) 
 ylabel(-0.2(0.2)0.4, labsize(vlarge) ) 
 xtitle ("Pre-Reform Men's Turnout (1918)", size(vlarge)) 
 xlabel(0(25)100, labsize(vlarge) ) 
 plotopts(msymbol(o) msize(huge)  lwidth(thick) lpattern(solid) color(black)) 
 ciopts(lpattern(dash) color(black) lwidth(medthick) msymbol (none)) recast(line) recastci(rconnected)  
 legend(off) scheme(s1mono) ysize(5) yline(0)  ;
#delimit cr
*sub-figure b
reg  ch_turnrat  Lturnoutm fact20 intel20 marw20 i.urban   i.PR_district, cluster(PR_district), if  year==1921
margins, at(Lturnoutm = (38 (1) 91))
#delimit ;
 marginsplot, 
 title("Parliamentary Elections", size(vlarge) ) 
 ytitle("Linear Prediction", size(vlarge) ) 
 ylabel(-0.2(0.2)0.4, labsize(vlarge) ) 
 xtitle ("Pre-Reform Men's Turnout (1918)", size(vlarge)) 
 xlabel(0(25)100, labsize(vlarge) ) 
 plotopts(msymbol(o) msize(huge)  lwidth(thick) lpattern(solid) color(black)) 
 ciopts(lpattern(dash) color(black) lwidth(medthick) msymbol (none)) recast(line) recastci(rconnected)  
 legend(off) scheme(s1mono) ysize(5) yline(0)  ;
#delimit cr





























